Introduction

Recent years have witnessed a growing interest in the characterization and study of the molecular organization of starches. This motivation is since the fine structure of starch plays an important role in the functional and nutritional features in starchy foods [1, 2]. For instance, the viscosity of starch gels and starch digestibility have a remarked molecular basis [3,4,5,6].

The development of chromatographic equipment and devices (columns and detectors) has been an important driver to achieve high-precision characterization of the fine structure of starch. HPSEC and HPAEC are mandatory chromatographic techniques to determine the CLD of amylopectin for the purpose of analyzing its structure–function and structure-digestibility relationship [7, 8]. Among the first studies on this topic is the work of Hizukuri [9] who used HPSEC to determine the CLD of isoamylase-debranched amylopectin, finding a polymodal distribution linked to the B-chains. On the other hand, Kasemsuwan et al. [10] reported that HPAEC separates individual amylopectin chains after debranching them from glucose to chains with DP values of around 70. Taking the above results as a reference, the CLD of amylopectin can generally be designated into four groups: DP between 6 and 12%, DP varying from 13 to 24%, DP ranging between 25 and 36%, and DP ≥ 37% [11]. Using this classification, starches from diverse botanical sources have been analyzed [11,12,13,14].

It is worth noting that starch dispersion is achieved experimentally, in principle by completely gelatinizing the starch in excess water by using a boiling water bath for 1 h, and then allowing the formed gel to reach room temperature. Afterwards, the debranching enzyme (isoamylase) is added to obtain the CLD of amylopectin either by HPSEC, HPAEC or fluorophore-assisted carbohydrate electrophoresis (FACE). In this framework, native and waxy starches are commonly well dispersed under boiling water conditions [15]. However, high-amylose starches show resistance to swelling and complete dispersion under boiling water conditions, thereby limiting the adequate determination of their CLD values, and therefore, an erroneous analysis about their fine structure can be provided [16]. In this same line, advances in this field have already been made. For example, Li et al. [17] achieved the complete gelatinization of high-amylose starches for the adequate analysis of their fine structure by dispersing the starch in dimethyl sulfoxide (DMSO) followed by precipitation with ethanol, and subsequent drying at 30  C for 48 h, to then be poured and dispersed in a boiling water bath for 20 min. with stirring. After this, Li et al. [17] allowed to cool to room temperature, and then debranched the starch using isoamylase. Vilaplana et al. [18] also achieved the complete gelatinization of high-amylose starches for their adequate characterization in terms of their fine structure by dispersing the samples in DMSO containing 0.5% w/w LiBr at 80 °C for 8 h with stirring at 350 rpm. However, these procedures are laborious and time-consuming.

Based on all the above, the hypothesis of this work lies in the assumption that water dispersion treatment by autoclaving high-amylose starches can be a fast and reliable method for obtaining adequate CLD values, in referred types of starches, compared to the water dispersion treatment under boiling water conditions, or contrasted with those of literature using DMSO (17, 18). The aim of the present study was to compare boiling water and autoclave treatments as previous procedures before enzymatic debranching, and subsequent quantification of CLD values in two commercial high-amylose corn starches (HYLON V and HYLON VII), for the proper analysis of its fine structure. All this was analyzed using two chromatographic techniques: HPSEC-RID and HPAEC-PAD.

Materials and methods

Materials

Two commercial high-amylose corn starches, HYLON V and HYLON VII, containing approx. 50 and 70% amylose according to the manufacturer (Ingredion, Tlanepantla, Estado de México, México) were analyzed. DMSO (HPLC grade, CAS: 67-68-5) and lithium bromide (LiBr, analytical grade, CAS: 7550-35-8) were purchased from Macron Fine Chemicals™ (Avantor, Gliwice, Poland) and Sigma Aldrich Chemical Co. (St. Louis, MO, USA), respectively. Isoamylase (EC 3.2.1.68) of Pseudomonas sp. with a specific activity of 180 U/mg (40 °C, pH 3.5) was bought from Megazyme (Wicklow, Ireland).

Water dispersion treatments: boiling water and autoclaving

For the water dispersion treatment by boiling water, 20 mg of each starch sample system was mixed with 6.4 mL of deionized water, and then the suspension was heated for 1 h in a boiling water bath with constant stirring. The water dispersion treatment by autoclaving was carried out as follows: Twenty mg of each starch sample system was dispersed in 6.4 mL of deionized water, and then autoclaved for 15 min at 120 °C in a pressure cooker. Subsequently, the starch samples were allowed to cool to room temperature and then debranched.

Debranching of the analyzed starch samples

After the water dispersion treatments, the starch samples were debranched separately with isoamylase employing the methodology reported by Hoyos-Leyva et al. [11], with some minor changes. Upon debranching, analyzed starch samples were lyophilized. In total, four sample systems were analyzed namely HYLON V and HYLON VII starches dispersed by boiling water, and HYLON V and HYLON VII starches dispersed by autoclaving.

Chromatographic analysis: HPSEC-RID and HPAEC-PAD

HPSEC-RID

For calibration, a series of pullulan standards (Polymer Standard Services, Mainz, Germany) with different molecular weights varying between 342 and 805,000 Da were used. For this, pullulan standards (4 mg) were dissolved in a DMSO/LiBr mixture (1.5 mL) at 80 °C (water bath) for 2 h with constant stirring at 350 rpm using a thermomixer (Benchmark, New Jersey, USA). Thereafter, the mixture was passed through a 0.45 μm syringe microfilter into the SEC vials. The HPSEC-RID system consisted of an Agilent 1100 chromatograph (Agilent Technologies, Waldbronn, Germany) coupled to a multi-angle laser light scattering (MALLS) detector and its complementary Optilab T-rEx dRI detector (DAWN® HELEOS II®, Wyatt Technology Co., Santa Barbara, CA, USA), which were set at 45 C. The chromatographic conditions were: DMSO/LiBr as mobile phase at a flow of 0.6 mL/min, injection volume 100 μL and oven temperature 80  C. The separation was carried out using a system of two analytical columns, GRAM 100 and GRAM 1000 (PSS, Mainz, Germany), connected in series. The refractive index increment (dn/dc) used was calculated by PSS (Mainz, Germany) following Vilaplana and Gilbert [19]: dn/dc = 0.0689 mL/g. All debranched and lyophilized starch sample systems (15 mg per sample system) were injected into the chromatograph exactly as the standards. The debranched SEC weight distribution (w(logVh,de)) vs. weight-average degree of polymerization (DP, Xde) was plotted, where debranched hydrodynamic volume (Vh,de) was determined using the Mark-Houwink equation:

$${V}_{h,de}=\frac{2}{5} \frac{K{M}^{1+\alpha }}{{N}_{A}}$$

NA represents the Avogadro number, K and α in the mobile phase (DMSO/LiBr) at 80 °C are 1.5 × 10−4 dL/g and 0. 743, respectively, for linear starch chains. To calculate M = 162.2 (X − 1) + 18.0, the molecular weights of the anhydroglucose monomeric unit (162.2 g/mol) and the water of the additional water in the final group (18.0 g/mol) were taken.

The amylose content in the starch sample systems analyzed was determined from the graph w(logVh,de) vs. DP, Xde. From this graph, the relationship between the area under the curve (AUC) of the branches with DP values greater than 100 and the AUC of the entire distribution of both the amylopectin and amylose populations was obtained. This way of quantifying amylose was validated by Fitzgerald et al. [20] and Vilaplana et al. [21].

HPAEC-PAD

The CLD of amylopectin of the different starch systems analyzed was carried out by HPAEC-PAD using the equipment and procedures previously described by us elsewhere [14].

Statistical analysis

Each sample system was replicated in duplicate, and the results were reported as the mean ± standard deviation (SD) using the ANOVA test, and then the mean of all pairwise groups (Tukey) was compared with a significance of p < 0.05 using the Sigma Plot v.12.0 software.

Results and discussion

The two high-amylose corn starches evaluated (HYLON V and HYLON VII) were better dispersed by autoclaving than by the boiling water dispersion treatment. This was evidenced by a better resolution of the HPSEC-RID elution profiles in both samples using said treatment (Fig. 1) even when debranched starch was dispersed in DMSO/LiBr before injection in the HPSEC-RID equipment. Thus, the autoclaving water dispersion treatment allows a better resolution of the starch fractions by HPSEC-RID, regardless of the type of high-amylose corn starch tested. The most marked improvements were observed for small elution volumes (up to 20 mL) linked to amylose and sparsely branched amylopectin. In this region, the elution profiles showed a well-defined curve without fluctuations, i.e., no interferences attributed to poorly dispersed starch fractions were evidenced. Therefore, the autoclaving water dispersion treatment allowed a better dispersion of high-amylose corn starches compared to the boiling water dispersion treatment.

Fig. 1
figure 1

High performance size-exclusion chromatography elution profiles of commercial debranched corn starches: a Hylon VII (red line) and b Hylon V (green line) dispersed by boiling water, and c Hylon VII (magenta line) and d Hylon V (blue line) dispersed by autoclaving

Full size image

Table 1 shows the mass fractions calculated as the area under the peaks of each fraction. The FI fraction related to the amylose content varied between 35.1 ± 3.5% and 43.3 ± 1.2% for HYLON V and HYLON VII starches dispersed by the boiling water treatment, respectively. Higher values of the FI fraction were reported by Martínez et al. [22] for HYLON V (55%) and HYLON VII (48.2%) dispersed in DMSO/LiBr and precipitated with ethanol, freeze-drying and dispersed in buffer under boiling water compared to those obtained here. FI fraction values were, however, increased in both starches dispersed by autoclaving compared to these same starches dispersed by the boiling water dispersion treatment (Table 1). This suggests that autoclaving water dispersion treatment of high-amylose corn starches increases the solubilization before debranching, resulting in a better analysis of their fine structure. Fractions II and III are associated with the long and short chains of amylopectin, respectively. In this regard, high-amylose corn starches normally have high ratios of long amylopectin chains to short amylopectin chains [23]. A similar trend was evidenced in this study. This fact was more evident for HYLON VII starch dispersed by autoclaving. An unclear effect on the ratio of fractions III (FIII—short amylopectin chains, A) to FII (FII—long amylopectin chains, B1, B2 and B3) (FIII/FII) between the starches dispersed by boiling water and their respective analogous starch dispersed by autoclaving was observed. Nevertheless, the highest value of the FIII/FII ratio was for HYLON V starch dispersed by autoclaving, i.e., HYLON V starch was the starch analyzed with the highest radio of short amylopectin chains to long amylopectin chains. This agrees with the SC/LC value obtained by HPAEC-PAD as will be analyzed later.

Table 1 High performance size-exclusion chromatography-refractive index detector (HPSEC-RID) of isoamylase-debranched high-amylose corn starches
Full size table

The debranched SEC weight distributions (w(logVh,de)) of the starch systems analyzed by HPSEC-RID displayed three peaks (Fig. 2): the first related to the short amylopectin chains (XAp1), the second attributed to the long amylopectin chains (XAp2) and the third associated with amylose chains (XAm) [24]. The DP values in the diverse fractions increased upon the autoclaving water dispersion treatment with respect to the boiling water dispersion treatment, regardless of the high-amylose corn starch tested (Table 2). Likewise, amylose content values determined with the Mark-Houwink equation increased in both starches tested when the autoclaving water dispersion treatment was used with respect to their respective analogous starch dispersed by boiling water (Table 2). The amylose content determined here is close to the values reported by the maker, 50 and 70% for HYLON V and HYLON VII, respectively. Therefore, the dispersion method by autoclaving and the calculation using the Mark-Houwink equation allow accurate determination of the amylose content and the DP of the amylose and amylopectin chains.

Fig. 2
figure 2

Debranched size-exclusion chromatography weight distributions (w(logVh,de)) of the starch systems analyzed: a Hylon VII (red line) and b Hylon V (green line) dispersed by boiling water, and c Hylon VII (magenta line) and d Hylon V (blue line) dispersed by autoclaving

Full size image
Table 2 Molecular parameters of debranched starches obtained from hi h performance size-exclusion chromatography-refractive index detector (HPSEC-RID)
Full size table

The CLD values of commercial high-amylose corn starches determined by HPAEC-PAD did not show a clear effect of the autoclaving water dispersion treatment compared to the boiling water dispersion treatment (Table 3, Fig. 3). No statistically significant differences (p ≥ 0.05) were observed in the ratio of short amylopectin chains to long amylopectin chains (SC/LC) for Hylon VII starch, i.e. the dispersion treatments used in this work did not affect the SC/LC ratio for Hylon VII starch. However, a statistically significant increase (p ≤ 0.05) in the SC/LC ratio was observed for Hylon V starch dispersed by autoclaving compared to Hylon V starch dispersed by boiling water. Even the Hylon V starch dispersed by autoclaving was the starch with the highest SC/LC ratio. This agrees with the results obtained by HPSEC-RID (highest value of the FIII/FII ratio). This suggests that autoclaving water dispersion treatment of Hylon V starch increased its debranching. However, lower SC/LC values were observed here compared to other studies: waxy corn starch (0.303) [25] and aroid starches (0.44) [11]. Therefore, the high-amylose corn starches analyzed had a low degree of amylopectin branching.

Table 3 Chain length distributions of commercial high-amylose corn starches by HPAEC-PAD
Full size table
Fig. 3
figure 3

Typical chain-length distribution (CLD) curves of amylopectin for of the starch systems analyzed: A HYLON VII and B HYLON V dispersed by boiling water, and C HYLON VII and D HYLON V dispersed by autoclaving

Full size image

Conclusions

The type of starch dispersion treatment affects the analysis of the fine structure of starch. High amylose corn starches dispersed by autoclaving showed more defined elution profiles by HPSEC-RID compared to their respective analogous starches dispersed by boiling water using the same quantification methodology. This resulted in a more accurate quantification of amylose content. Thus, a better quantification of amylose content could be suggested by autoclaving water dispersion treatment, since amylose content may be underestimated in high-amylose starches dispersed in boiling water. Regarding the analysis of the structure of amylopectin, the autoclaving water dispersion treatment was initially hypothesized as a treatment that would, in general, improve the debranching of the starch, and as a result, a better analysis of the fine structure of the starch would be achieved. Nonetheless, the results obtained here demonstrated that depending on the type of dispersion used, the values of the amylaceous fractions are altered. Hence, this affects the proper interpretation of the fine structure of at least high-amylose corn starches.